测试大模型在无法回答的问题上是否能拒绝作答，发现越强的模型越不会主动放弃。
AbstentionBench: Reasoning LLMs Fail on Unanswerable Questions
- 构建20个数据集组成的基准，涵盖无答案、信息不足等10类难问题
- 20个前沿大模型平均拒答率下降24%，推理训练反而更糟
- 提示词可临时提升拒答，但无法解决模型本质缺陷
为确保大语言模型在日常和高风险场景中可靠部署，识别何时不应回答与正确回答同样重要。真实用户提问常存在信息不全、逻辑错误或根本无解等问题，要求模型具备不确定性推理能力并选择性拒答。然而，拒答机制仍研究不足，缺乏系统评估框架。本文提出AbstentionBench，一个覆盖20个多样化数据集的大规模基准，包含未知答案、信息不全、虚假前提、主观解释及过时信息等问题。评估20个前沿大模型发现，拒答仍是未解难题，模型规模扩大对此帮助甚微。尽管近期推理型大模型在复杂任务上表现优异，但意外发现推理微调使拒答能力平均下降24%，即使在数学与科学领域也如此。我们还发现，精心设计的系统提示虽可提升实际拒答率，却无法改变模型对不确定性的根本认知缺陷。我们开源AbstentionBench以推动大模型可靠性研究。
原文摘要 · Abstract (English)
For Large Language Models (LLMs) to be reliably deployed in both everyday and high-stakes domains, knowing when not to answer is equally critical as answering correctly. Real-world user queries, which can be underspecified, ill-posed, or fundamentally unanswerable, require LLMs to reason about uncertainty and selectively abstain -- i.e., refuse to answer definitively. However, abstention remains understudied, without a systematic evaluation framework for modern LLMs. In this work, we introduce AbstentionBench, a large-scale benchmark for holistically evaluating abstention across 20 diverse datasets, including questions with unknown answers, underspecification, false premises, subjective interpretations, and outdated information. Evaluating 20 frontier LLMs reveals abstention is an unsolved problem, and one where scaling models is of little use. While recent reasoning LLMs have shown impressive results in complex problem solving, surprisingly, we find that reasoning fine-tuning degrades abstention (by $24\%$ on average), even for math and science domains on which reasoning models are explicitly trained. We find that while a carefully crafted system prompt can boost abstention in practice, it does not resolve models' fundamental inability to reason about uncertainty. We release AbstentionBench to foster research into advancing LLM reliability.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。